Secure and reliable computing remains an open problem. Just in the past year, security vulnerabilities continued to be discovered in widely-used applications, and remote attacks that take advantage of these vulnerabilities can cause an enormous amount of damage. Runtime monitoring is a promising approach for addressing a wide range of security and reliability problems in a general and transparent fashion. However, software-only approaches have high performance overheads on applications, and using custom hardware to mitigate these overheads can result in rigid schemes that are unable to adapt to new threats. In this thesis, we will describe steps to build an architectural framework for runtime monitoring that can address the drawbacks of previously proposed runtime monitoring schemes by utilizing reconfigurable hardware and optimizing for performance.